Fgf6, a member of the Fibroblast Growth Factor (FGF) family, is developmentally regulated and its expression is highly restricted in the adult. To gain further insight into the role of Fgf6, we studied its expression during embryogenesis using RNA in situ hybridization. Fgf6 expression is restricted to developing skeletal muscle. Fgf6 transcripts are first detected in the somites at 9.5 days post-conceptus, and expression continues in developing skeletal muscles up to at least 16.5 days post-conceptus. Fgfr4 is a putative receptor for FGF6. Its pattern of expression during myogenesis overlaps that of Fgf6, but both genes are not expressed in exactly the same population of cells. In addition, recombinant FGF6 protein is able to repress the terminal differentiation of myoblasts in culture, providing additional support to the concept that FGF6 plays an important role in myogenesis.

Using isogene specific probes and in situ hybridization on sections, we have examined the expression of structural and regulatory genes in the mouse embryo during the formation of cardiac and skeletal muscle. The temporal and spatial information thus obtained about the onset of expression of muscle genes provides insight into the regulation of myogenesis in vivo. Actin and myosin sequences present in different compartments of the adult heart are initially all co-expressed in the cardiac tube (between 7-8 days). The process of spatial restriction to atrial or ventricular compartments of the heart takes place asynchronously later. In contrast, the onset of expression of actin and myosin genes in the first skeletal muscle, the myotome, which corresponds to the central compartment of the somite, as well as their subsequent down-regulation in different skeletal muscle masses, takes place very asynchronously. One might predict that factor(s) responsible for the transcriptional activation of these genes are present in sufficient quantity in the cardiac tube, whereas in skeletal muscle individual genes are responding to variable levels of factor(s). In fact the four myogenic regulatory sequences present in the mouse - MyoD1, myogenin, myf-5 and myf-6 - do show distinct patterns of expression during the development of skeletal muscle. None of these sequences have been detected in the heart. In the myotome there is no general correlation between the appearance of a particular myogenic sequence and the activation of a particular structural gene. A striking example of this is provided by the muscle isoform of creatine phosphokinase. We would propose that each muscle structural gene has a different threshold of activation, depending on the quantity and nature of the myogenic factor present. We have also examined the onset of expression of the X-linked dystrophin gene known to be expressed in adult heart and skeletal muscle. In the myotome dystrophin transcripts are first detected at the time when myosin heavy chains first accumulate and muscular contraction is initiated. In contrast in the cardiac tube dystrophin transcripts are not detected initially, at a time (from 8 days) when the heart contracts. This observation can be correlated with the pathology of the disease which points to a more essential role of dystrophin in skeletal muscle. No muscle structural gene examined is expressed in the somite prior to myotome formation. If the myogenic regulatory sequences are implicated in muscle cell determination then they should be expressed in the dermomyotome of the immature somite which gives rise to muscle precursor cells.(ABSTRACT TRUNCATED AT 400 WORDS)

The first striated muscle to form during mouse embryogenesis is theheart followed by skeletal muscle which is derived from the somites. The expressionof genes encoding muscle structural proteins and myogenic regulatory sequencesof the MyoD1 family has been examined using 35S-labelled riboprobes. In thecardiac tube, actin and myosin genes are expressed together from an early stage,whereas in the myotome, the earliest skeletal muscle, they are activatedasynchronously over days. They are not expressed in the somite prior to myotomeformation. One potential muscle marker, carbonic anhydrase III, is expressed inearly mesoderm and subsequently in the notochord, similarly to the Brachyurygene. The myogenic sequences are not detectable in the heart. In the myotomethey show distinct patterns of expression; this is discussed in the context of theirrole as muscle transcription factors. myf-5 is the only myogenic factor sequencepresent in the somite prior to muscle formation and thus is potentially involvedin an earlier step of muscle determination. It is also present in the early limb bud,but the status of myogenic precursor cells in the limb in this context is less clear.

We define the spatial and temporal patterns of expression of the gene encoding the glycolytic enzyme, β-enolase, during mouse ontogenesis. Transcripts were detected by in situ hybridization using 35S labelled cRNA probes. The β-enolase gene is expressed only in striated muscles. It is first detected in the embryo, in the cardiac tube and in newly formed myotomes. In the muscle masses of the limb, β gene expression occurs at a low level in primary fibers, and subsequently greatly increases at a time which corresponds to the onset of innervation and secondary fiber formation. Later in development, it becomes undetectable in slow-twitch fibers. Our results demonstrate the multistep regulation of the β-enolase gene. The regulation of this muscle-specific gene in somites is discussed in terms of the myogenic sequences of the MyoD family shown to be present when it is activated.

HNF1 (Hepatic Nuclear Factor 1) and vHNF1 are transcriptional regulators containing a highly divergent homeodomain. The first was initially found in liver nuclear extracts and is crucial for the transcription of albumin and many other hepatocyte specific genes, while the second was found in dedifferentiated hepatoma cells. Both recognize the same DNA binding site and can form homo and heterodimers in vitro and in vivo. In situ hybridization analyses have been performed to delineate the spatial and temporal pattern of expression of vHNF1 relative to HNF1 during mouse embryogenesis. The results show that accumulation of vHNF1 mRNAs expression is detected in several tissues of the embryo of both endodermal and mesodermal origin. Expression occurs in the yolk sac, the primitive gut, the liver primordium, and at different stages of kidney development in polarized epithelial structures and usually precedes that of HNF1. vHNF1 expression seems particularly prevalent with morphogenetic events in the kidney and may be a marker for certain polarized epithelium.

We have analysed by in situ hybridization the expression of myf-5, the murine homologue of the human myogenic regulatory sequence myf5, during embryogenesis in the mouse. myf-5 sequences were first detected in the earliest somites (from about 8 days p.c.) in the dermomyotome, before formation of the dermatome, myotome and sclerotome. The dermomyotome is classically considered to give rise to the precursor muscle cells of body and limb skeletal muscle. myf-5-positive cells were also detected early in the visceral arches and limb buds. In this case, as in somites, myf-5 expression precedes that of the two related myogenic regulatory sequences, myogenin and MyoD1, and indeed any other skeletal muscle marker examined to date. myf-5 is not detected at any stage in developing cardiac muscle. From 11.5 days p.c., the level of myf-5 transcripts begins to decrease to become undetectable (by in situ hybridization) from 14 days p.c. Both the appearance and disappearance of myf-5 follow the anteroposterior gradient of somite formation and maturation in the embryo. The time and place of myf-5 expression are consistent with a role in the early events of myogenic differentiation, possibly during determination of the myogenic lineage.

The 150-base-pairs region located upstream of the transcriptional start site of the rat albumin gene contains all of the critical sequences necessary for this gene's tissue-specific expression in rat hepatoma cells. In transient expression assays using an improved CAT system or direct mRNA analysis we were able to detect a faithful transcription from the albumin promoter in albumin-negative dedifferentiated H5 hepatoma cells which was 250-fold weaker than in differentiated H4II hepatoma cells producing albumin. This strong tissue specificity could be completely overcome through the cis action of a non-tissue-specific enhancer. Two upstream regions from nucleotides -151 to -119 and from -118 to -94, were required for efficient transcription in H4II cells. Each region contained a sequence motif highly conserved among different species. The effect of the -151/-119 region was strictly tissue specific, while the -118/-94 region was also involved in the low level of transcription observed in H5 cells. Finally, sequences between the CCAAT box and the TATA box also contributed to the overall tissue specificity of rat albumin gene transcription.

Numerous studies of cell hybrids have indicated that somatic cells produce negative regulators (extinguishers) that prevent the expression of functions foreign to their own differentiation. Here, we report genetic evidence of such control. In microcell hybrids between well-differentiated rat hepatoma cells and microcells of mouse fibroblast L cells, the extinction of albumin synthesis is directly related to the presence of a single specific chromosome of the mouse fibroblast parent. The expression of several other hepatic functions is not affected. Transfection of these hybrids with a recombinant plasmid, containing a tissue-specific control element of the upstream region of the rat albumin gene linked to coding sequences of the chloramphenicol acetyltransferase gene, reveals that extinction acts on or via this cis-control element.

We have previously identified an Msp I site at the 5' end of the rat albumin gene whose undermethylation is necessary but not sufficient for stable albumin expression in rat hepatoma cells. We have also shown that the extinction of albumin expression in somatic hybrids is not the result of methylation at this site, since for two different crosses, rapid extinction was found to occur in the absence of any de novo methylation of the previously active gene. In the present study, we examine albumin expression and albumin gene methylation for independent hybrid clones isolated from crosses between albumin expressing rat hepatoma cells and cells of two different non-expressing lines. The cells from hybrid clones of both crosses are characterized by stable extinction of albumin expression. Moreover, we find that de novo methylation of the "extinguished" albumin gene can occur in somatic hybrids, but only some weeks after the gene has ceased to be expressed.

We have constructed a transient expression vector containing 400 bp of rat albumin gene immediate 5′‐flanking sequences inserted 5′ to the bacterial enzyme chloramphenicol acetyl transferase (CAT). We have transfected various clones of rat hepatoma cells representing different states of expression of the liver phenotype with this vector (pALB‐cat) and also with two control vectors containing viral promoters (pSVE‐cat and pRSV‐cat), and measured activity of the bacterial enzyme CAT in cellular extracts 48 h later. The albumin flanking sequences are able to direct highly efficient CAT expression, compared with the control vectors, only in cells which express their own albumin gene: the albumin‐negative hepatoma cells are at least 100 times less efficient in expressing CAT after transfection with the pALB‐cat plasmid than are the albumin‐positive ones. An unexpected result of our study is the total inability of the rat albumin flanking sequences to direct expression in albumin‐producing mouse hepatoma cells.

In earlier work we identified at the 5' end of the rat albumin gene an Msp I site whose undermethylation appears to be necessary but not sufficient for stable expression of the gene in rat hepatoma cells. Here, we ask whether the block to expression of albumin production, which occurs when rat hepatoma cells are hybridized with cells that do not produce the protein, could be the result of de novo methylation of this site. In two types of somatic hybrids, rat hepatoma-mouse L cell fibroblasts, and rat hepatoma-dedifferentiated variant rat hepatoma cells, extinction occurs and is maintained during the first 5-15 generations after fusion. During this time the Msp I site of the now inactive rat albumin gene remained unmethylated.

We have measured methylation of the albumin gene in clones of rat hepatoma cells that vary quantitatively in their rates of synthesis of albumin and in variant and hybrid cells that produce no albumin. Although the albumin gene is undermethylated for its entire length in rat liver, only the 5′ end is ever undermethylated in hepatoma cells. Moreover, undermethylation of the 5′ end of the gene appears to be necessary for stable expression of the albumin gene in hepatoma cells. Since undermethylation of this region is found in some variant cells that fail to produce albumin, it is not a sufficient condition for albumin gene expression. Despite the excellent correlation between undermethylation of the 5′ end of the albumin gene and its stable expression, the results argue against the possibility that the methylated state of such genes during development determines whether they will or will not be expressed.

Rat hepatoma clones whose cells do and do not produce albumin, as well as somatic hybrid between the two types of cells, have been examined for albumin mRNA. A direct proportionality between the rate of albumin production and the concentration of albumin mRNA sequences was found for all albumin-producing hepatoma and hybrid clones, indicating that rate of synthesis of the protein is determined by the concentration of its mRNA. Albumin-negative dedifferentiated variant and somatic hybrid cells contain fewer than one to five molecules of albumin mRNA per cell; the block in expression of the gene appears to be at the same (probably transcriptional) level in variants and their somatic hybrids.